Beacon — Startup Management & Decision Workspace

Early-stage founders and small engineering teams struggle with context fragmentation across fragmented tools—spreadsheets for runway, Kanban boards for sprint tasks, and disparate chat threads for critical decisions. There is no single workspace connecting daily execution directly to financial health and AI-assisted strategic advice.

Engineered Beacon to provide a single, cohesive workspace that keeps founders and engineering leads grounded in operational reality by uniting tasks, ledger transactions, and a Gemini-powered AI co-founder injected with live workspace context.

Next.js App Router full-stack architecture with server-rendered workspace views, API route handlers connecting to MongoDB for structured documents (tasks, transactions, team rosters, decisions), Socket.IO channels for real-time collaboration updates, and a Gemini 2.5 Flash API integration acting as a conversational AI co-founder.

• Engineered unified workspace data models connecting sprint tasks, financial ledger items, team members, and decision logs
• Implemented Google OAuth authentication and session management for secure multi-user workspace isolation
• Constructed structured prompt schemas that inject current startup metrics (burn rate, milestone progress) into Gemini API requests
• Created interactive financial runway visualizers and Kanban boards with optimistic UI updates

export async function buildAdvisorContext(workspaceId: string, userQuery: string) {
  const [tasks, ledger, metrics] = await Promise.all([
    TaskModel.find({ workspaceId, status: { $ne: 'completed' } }).limit(20),
    LedgerModel.find({ workspaceId }).sort({ date: -1 }).limit(30),
    getWorkspaceRunwayMetrics(workspaceId)
  ]);

  const systemPrompt = `You are an AI co-founder advisory engine. Context: Current Runway: ${metrics.runwayMonths} months. Monthly Burn: $${metrics.monthlyBurn}. Active Tasks: ${tasks.length}.`;
  return geminiClient.generateContent({ systemPrompt, prompt: userQuery });
}